LLM Reference

Qwen2.5-72B

Released
2025-10-10
Last refreshed
2026-04-15
Status
Researched 185d ago
Open sourceCommercial use: permittedRAGAgentsLong contextClassificationJSON / Tool use

Qwen2.5-72B is a released rag, agents, and long context model with open-source and 128k context; evaluate it while provider pricing coverage matures.

Use it for

  • Teams evaluating rag, agents, and long context
  • Workloads that can use a 128k context window

Do not use it for

  • Cost-sensitive launches that need sourced token pricing
  • Vision or document-understanding workloads
  • Teams that need a tracked hosted API route today
Specifications
Family
Qwen2.5
Released
2025-10-10
Context
128k
Parameters
72B
Knowledge cutoff
2024-09
Openness
Open source
License
Apache 2.0OSI-approvedCommercial use: permitted
Weights
Available
Code
Unknown
Created by
AlibabaAlibaba

AI research institute of Alibaba Group.

Hangzhou, Zhejiang, China
Founded 2017
Website
Pricing

No tracked provider token pricing is available yet.

About

Alibaba Qwen2.5 72B model with improved reasoning and 128K context window. Open-weights, optimized for diverse tasks including coding and analysis.

Qwen2.5-72B is an open-source model in the Qwen2.5 family. The structured metadata tracks a 128k-token context window, function calling, and tool use. Headline tracked benchmarks include MMLU PRO 72.0.

Top use-case fit: coding, agents, and build tasks

RAG

Included by capability and metadata signals in the decision map.

Agents

Included by capability and metadata signals in the decision map.

Long context

Included by capability and metadata signals in the decision map.

Provider price ladder

No tracked provider token pricing is available for this model yet.

Capabilities

Function CallingTool UseFine-tuning

Benchmark peer barsfor Classification

MMLU PRORank 69 of 105
Gemini 3 Pro
91.8
Gemini 3.1 Pro Preview
91.0
Qwen3.7-Max
89.6
Claude Opus 4.6
89.1
Qwen2.5-72Bcurrent
72.0

Benchmark scores(1)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
MMLU PRO72.0—https://huggingface.co/spaces/TIGER-Lab/MMLU-Pro

Migration checks

No linked migration route is available for this model yet.

Compare Qwen2.5-72B with other models

Show all 22 popular comparisonssorted by 7-day search impressions
Qwen2.5-72B vs Llama 3 8B Instruct40Qwen2.5-72B vs GPT-5.5 Instant39Qwen2.5-72B vs Claude Sonnet 4.533Qwen2.5-72B vs Claude Opus 4.626Qwen2.5-72B vs Grok Build 0.119Qwen2.5-72B vs Claude Opus 4.514Qwen2.5-72B vs GLM-5 9B14Qwen2.5-72B vs GLM-5 Turbo12Qwen2.5-72B vs Code Cushman 00212Qwen2.5-72B vs Mistral Large 212Qwen2.5-72B vs Claude Haiku 4.512Qwen2.5-72B vs o3 Deep Research8Qwen2.5-72B vs Grok 4.38Qwen2.5-72B vs Kimi K2.58Qwen2.5-72B vs GPT-5.27Qwen2.5-72B vs GPT-5.47Qwen2.5-72B vs GLM-5V-Turbo7Qwen2.5-72B vs GPT-5.4-Cyber5Qwen2.5-72B vs Gemini 2.5 Flash Live API4Qwen2.5-72B vs Mistral Large 2 (2407)3Qwen2.5-72B vs Mistral Large 3 675B Instruct3Qwen2.5-72B vs Code Davinci 0012

Frequently asked questions

What is the context window of Qwen2.5-72B?

Qwen2.5-72B has a context window of 128k tokens.

When was Qwen2.5-72B released?

Qwen2.5-72B was released on 2025-10-10.

What benchmarks has Qwen2.5-72B been tested on?

Qwen2.5-72B has been evaluated on 1 benchmark, including MMLU PRO.